Sunday, 27 March 2022

contra aftab contra gipps contra seth

In a recent blogpost Awais Aftab has expressed some welcome disagreements with what I wrote in my critique of Anil Seth's conception of perception 'Seth's vision' post. What follows is my response.

heliocentric or geocentric: what's the truth?

Aftab writes of a 

scientific model according to which [the] sun is in orbit around the earth. This scientific model is decidedly false; the sun is not in orbit around the earth, however things appear to us.

Here Aftab's flatly disagreeing with what I wrote in my original post - I expressly denied that this is a decidedly false claim. But he doesn't argue his case - so I'll now just reiterate and expand mine. It's important to note, before we begin, that whilst the explicit topic is astronomical, the real issue - to which we'll return after the physics is expounded - concerns instead the relevance of considerations of linguistic meaning when what's under discussion are allegedly scientific questions. 

So: I take it as axiomatic that if we're to model the movements of bodies we must always first stipulate a reference frame. Being decidedly terrestrial creatures it's our typical habit to use the earth's surface as this 'immobile'-by-definition reference frame for ordinary modelling and measurements. This use amounts to a tacit stipulation, a stipulation which gives us a rule of representation: movements of people and cars and birds are to be measured relative to the earth's surface (and not, say, relative to that hypomanic rabbit running hither and thither over there, since that'd really complicate everything terribly). Such rules of representation aren't themselves representations: they aren't, that is, themselves happily styled 'true' or 'false'. There's no such thing as a 'correct stipulation'. There are just reference frames which are more or less useful depending on our needs. 

Now, when we're discussing the movements of celestial bodies, we might surely offer either the sun or the earth as our reference frame. We could of course stipulate the moon but, depending on the scientific context, that'd probably make the maths very complex. Might we follow Newton and pin our frame instead to the 'fixed stars'? Well, we'd first need to say in what sense they're 'fixed', since it's generally thought that, what with the  expansion of the universe, many of them are moving apart from one another. And we'd surely want to avoid Newton's notions of 'absolutely fixed', or 'absolute motion', or 'absolute duration', etc., since, so far as I know, that idea - of motion or duration not relative to any particular spatial or temporal reference frame but just, somehow, 'in itself' - has never been provided with a meaning. But at any rate, if we stipulate the earth, then the sun will properly be said to be moving around it - and vice versa. 

Given this I simply find it hard to know what's meant by Aftab's 'This scientific model is decidedly false; the sun is not in orbit around the earth, however things appear to us.' But perhaps he would say 'But in science we always call the simplest model decidedly true and the more complex model decidedly false.' Well, we could say this - though I don't think we always, or even usually, do. (I think we typically just talk of the advantages of simplicity, and leave off talk of truth and falsity for when instead we're talking about representational fit. That, at least, stops us getting into unnecessary muddles.) But, sure, I'm happy if we do decide to use the word 'true' like that: now we all know what we mean, philosophy has done its clarificatory job, and we're back not talking past each other. So now, because of what it does for astronomical science, we'll say that a heliocentric model is there the 'true one'. Or because of what it does for botanical science (you're trying to measure the light shifting over a plant's leaves, and it's rather easier to think of the sun moving overhead rather than of the plant orbiting the sun), we'll say that the geocentric model is there the 'true one'. My concern though is that there was rather a whiff of 'but in fact, or really, or as science reveals to us, it's the earth that orbits the sun and not vice versa' to both Aftab's and Seth's presentations. And, at least on its face, such an 'in fact' appears to rather transcend mere matters of simplicity. Instead it seems to be an 'in fact' which has (...in fact...) rather gone on semantic holiday. An 'in fact' which, like Newton's notions of 'absolute' motion or duration, has attempted to prescind from particular contexts of enquiry, forgetting that it's only ever relative to a context that such notions enjoy a meaning.

Well, wait... why does any of this matter - wasn't the real topic the cognitive neuroscience of perception? So, well, yes: this all has to do with how we understand the significance of philosophy - in particular: philosophical reflection on linguistic meaning - for (would-be) scientific enquiry. So let's now turn to this.

language... or science?

After claiming that the geocentric view is 'decidedly false' Aftab goes on to urge that:

We shouldn't be so preoccupied with the language itself that we forget there is an independent scientific question to be asked. In a similar way, questions pertaining to the use of language about perception should not lead us to ignore the scientific questions at hand.

Similar comments regarding the relation of linguistic philosophy to matters scientific are offered later:

To remember how to correctly use the word perceive doesn’t by itself tell us what the correct scientific psychological account of perception is, and any scientific psychological account of perception has to take into account the fact that the brain is confined inside the skull and only has access to signals in the sensory nerves. 

Does the brain approximate Bayes’s rule in the process of perception? This is an empirical question, to be settled by scientific inquiry, but it certainly cannot be settled or eliminated by an analysis of ordinary language.

So, first, what is the 'independent scientific question to be asked' here, in this first context which, to recall, had to do with celestial movements? I confess the only one I can think of as a candidate might go something like: 'But, really, Dr Gipps, please leave aside all these parochial matters to do with this or that human-interest-relative context, and turn instead to matters of the raw scientific truth: just tell us, Dr Gipps: does the earth orbit the sun, or does the sun orbit the earth?!!' Yet, as I was at pains to show, it's here precisely philosophical considerations regarding language use - i.e. considerations regarding what it makes sense to say - and especially of distinguishing rules of representation (which as such aren't usefully described as true or false) from representations (which as such are either true or false) - which reveals that this just isn't a good question! We might at first think we know what it means. (Just like Anscombe, at first, thought she knew what she meant when she said that it seemed to her as if the sun went round the earth.) But then we think it through, and it turns out that we don't!

Turning now to perception, Aftab is of course correct that recalling how the word 'perceive' is properly used tells us nothing about which bona fide scientific theory of perception is correct. (And nothing in my original post declared anything else.) Where linguistic philosophy comes into its own, though, is when it helps clarify whether we are indeed asking bona fide scientific questions. My claim was that Seth seemed to think he was asking and answering clear scientific questions... but that this may be illusory. Now, philosophy can't help us answer scientific questions - but it can help us distinguish genuine from merely ersatz such questions. So, just as Seth and Aftab seem - if I'm not mistaken - to wrongly take 'Does the sun orbit the earth, or is it that the earth orbits the sun?' as a straightforwardly clear question - so too does Seth, as I read him, take his particular questions and answers regarding 'how we perceive' to enjoy a clear sense. But the appearance of meaningfulness, I claim, may be only a product of certain prior misunderstandings regarding how concepts work. If you conceptually assimilate donkeys to positive numbers, you may well think it a straightforwardly intelligible scientific question to ask just what the square root of a particular donkey is. (I don't say that we couldn't give this question a sense; only that it doesn't enjoy one on its face.) To undo that assimilation is not to obstruct scientific progress, but rather to point out that it seems we didn't so much as have a scientific question on the table. From Gellner's idiotic Words and Things onwards, philosophers have accused linguistic philosophy as being somehow antithetical to science. As I see it, however, this is entirely backwards. No science at all can happen if the questions being asked are not even empirical because they're riddled with conceptual confusion. By helpfully picking this apart, the linguistic philosopher is the one contributing to strengthening the scientific framework, whereas the would-be scientist who just ploughs on regardless despite their tacit confusion is holding back scientific progress by instead covertly indulging metaphysical nonsense.

brains and their owners

Aftab agrees that there is 'no mystery to how I see thing. I just do. I look around and see the world in all its beauty and ugliness. But in order for the brain to make this possible, an explanation is needed.' Brains however made perception possible long before explanations were available; I suspect Aftab really means: 'in order for us to understand how brain activity makes perception possible, we require a scientific explanation'. This to me seems utterly unobjectionable: sure, explanations aren't always required before we can understand something, but when we've got 'how does that work?' questions going on, an explanation will be just the ticket! So, yes: so far, so unobjectionable. Yet after noting that I suggest that an answer in neurological (I think 'neuroscientific' would have been a happier word choice by me) terms is required here, Aftab urges that this would:

not offer us the explanation we need; what will be missing will be explanation that connects the neurological activity to the perceiving I. There is, therefore, a cognitive and psychological question here as well. What cognitive and psychological processes are involved in our ordinary experience of perception?

I confess to not yet understanding this. If I want to know how, say, a Porsche can accelerate so quickly - the acceleration admittedly being a property of the whole car - I'll naturally be satisfied with an answer in purely mechanical terms, one which tells me about the functioning of the carburettor etc. Sure, I'll want to know too how the carburettor is connected to the throttle and the fuel supply and the engine and thereby to the wheels - but there seems to me no requirement for an extra kind of story, told at some other level of explanation, wherein I relate all of this to the accelerating car. Or, well, perhaps other details will be relevant too: depending on where I was coming from when I asked about the Porsche's acceleration, I might find it more illuminating to hear about the extraordinary funding of their R&D relative to other car manufacturers. But at any rate, we can I think imagine someone who is only interested in the question of what it is in the car itself that enables its rapid acceleration. And here they will, I think, be satisfied with a story about the mechanical goings on under the bonnet. So too a story about the eye, the optic nerve, the striate cortex, etc etc., which outlined the machinery of perception and detailed its operation could, I think, work in a similar way. 

Aftab goes on to tell us that:

any scientific psychological account of perception has to take into account the fact that the brain is confined inside the skull and only has access to signals in the sensory nerves. The relationship between the voltage changes in the nerve membranes and the world outside the sensory organs is not a question that has a straightforward obvious answer, and I refuse to accept that this is a meaningless question that arises only because we are confused about how we use the word “perceive”!

Now I'm not sure what it means to talk of a brain being 'confined' (perhaps 'safely contained'?) in a skull, and of it only having 'access' to this or that sensory signal. But perhaps the Porsche analogy can help us here. So you notice that whilst I wanted to understand the acceleration of the whole car, we somehow talked mainly about the carburettor and other internal components. The carburettor, however, doesn't have direct 'access' to the wheels or the road; it only has only has 'access' to the fuel and the air. This, you suggest, makes for an explanatory problem. ... Well, I demur. It may well be that the relation of acceleration to air intake is not straightforward: much will also depend on the current pitch of the road, the current speed, the drag of the vehicle etc. By all means, let's look at the whole picture. And the same - except more so - will be true of the 'relationship between the voltage changes in the nerve membranes and the world outside the sensory organs'. None of these enquiries will involve us in asking meaningless questions (and, note, I never claimed that they did); none of them presuppose confusion about how the word 'perceive' is used. What would be a confusion, though, would be the assumption that the brain's generation of perception / the carburettor's generation of acceleration required it to reconstruct or represent or make inferences regarding what's happening outside the skull / engine. (Or to put it otherwise: what would be a confusion would be if we conflated 'access' in the mechanical sense of i) enjoying this or that degree of causal connectedness to the world, with 'access' in the epistemic sense of ii) making the world's acquaintance.) No: such activities, if we're meaning them in their normal senses, are properties not of the brain or carburettor but instead, at times, of the person whose brain or car is under discussion.

prediction and inference

Let's now turn to that last question about brainy inferences etc in more detail. The part of Aftab's critique I find most thought-provoking is this: 

it strikes me as quite valid to hypothesize or talk about prediction in an analogous way to ordinary language but in a manner that doesn’t require intention or agency, etc. This is especially because we can meaningfully talk about mathematical models making predictions, and a variety of non-intentional, non-agential systems can enact said mathematical models.

Johannes Kepler
Now, I confess that while I think I can understand what it is for a mathematical model to make a prediction, I'm not entirely sure what it means for a system to enact a mathematical model, and I'm not at all sure what it means to say that the modelled system is itself making predictions. ... But let's turn again to the celestial analogue for a first pass. So: Kepler provided us with an equation to figure out the movements of two celestial bodies relative to one another. . We can model the moon's movements around the earth using this equation. If we said that 'the moon-earth system enacts this model', we're basically saying the same thing in different words. And we can use the model to help us predict the moon's location. But note, we can now deploy metonymy to arrive at: the model predicts where the moon will be in an hour.

So, if that's what Aftab means by a 'system enacting a mathematical model', then I should say that I can see nothing wrong with it. To transfer it now to the context of the brain, we might say this: Neuroscientists can model, and predict, how activity in part of the brain covaries with past and present sensory stimulation and motor activity (say) using a particular model. We then shunt the words around a bit to arrive at: 'this is the model the brain enacts'. We will say the same too for the release of insulin in the pancreas in response to blood sugar changes etc. We can model this mathematically; lo: the pancreas enacts the model. But what we don't yet get to, from the notion of a mathematical model making predictions, is a cogent sense in which the planets, the pancreas, or the brain are themselves predicting anything. So it looks to me like we'll need something other than the metonymic, conceptually parasitic, notion of 'mathematical models making predictions' to get us to a sense for 'the brain makes predictions'. 

linguistic innovation

This takes me to a part of Aftab's critique which I thought simply unfair:

Helmholtz, when he described perception as an inference, used inference in an analogous way: “[the “psychical activities” leading to perception] are in general not conscious, but rather unconscious. In their outcomes they are like inferences insofar as we from the observed effect on our senses arrive at an idea of the cause of this effect.” (Helmholtz 1867) [my emphasis, notice use of “like inference” suggesting an analogy].

The... problem I have is with the insistence that unless an explicit definition is offered, the use must be considered muddled or nonsensical. Just because Helmholtz does not further specify what “like inference” is supposed to be, does that make it muddled and nonsensical? I don’t think so. Scientific ideas often begin as a sort of analogy, and are further refined and made more precise over time. Didn’t Wittgenstein have something to say about the meaning of a word being its use? If a word is being used by an entire community of scientists, can we not recognize that use as legitimate, even if a formal definition is lacking?

I leave aside the issue of whether Helmholtz or we know what Helmholtz is saying when he talks of
Hermann von Helmholtz

unconscious psychical activity leading to a perception, or of perception involving us ('we') working out what the causes of sensory stimulations are. I happen to find that all exceptionally murky, but it's not what Aftab is after. Instead he's pointing to the idea that Helmholtz's inferences are only like ordinary inferences. So, ok: fine! But then Aftab effectively claims that I was saying that unless Helmholtz, Seth, whoever, offers an explicit definition of a term used in a new sense, that new use must be considered nonsensical or muddled. But this really wasn't at all what I was saying - and I also don't think I've seen any Wittgensteinian commentators on cognitive neuroscience say anything like this. 

The easiest way to get clear about this is to rehearse the dialectic. As I offered it, my challenge was two-pronged, and went like this:

a) If you, cognitive neuroscientist, are intending such and such a term in its ordinary sense, then your claims are, for reasons I've given, simply nonsensical.

b) If however you're using it in a new, or somewhat new, sense, a sense that we readers don't yet know, might you please both acknowledge when this is so (so we don't wrongly assume that you mean it in the ordinary sense), and also make this new meaning clear for us? If it's a new technical scientific sense, then yes a definition really would be lovely, since that's rather how we tend to do things in science. However perhaps you might instead give some paradigmatic examples, or do what you can to make clear what would falsify your claims, or elucidate whatever ascription conditions you can for the concept, etc.? Doing literally nothing to make yourself clear, however, well, dude: that's not cool! For making oneself clear isn't just a job for philosophers! It's part and parcel of ordinary responsible scientific practice.

This is why, in my original post, I kept asking if Seth would tell us what he meant, rather than simply said 'what you're saying is nonsense'. I think I demonstrated that it would be nonsense if he was using the terms in the ordinary way. I think I demonstrated that Seth is bizarrely insouciant regarding the need to tell a reader of a popular book what his terms mean - and I do rather suspect that this might be because he doesn't think he is using terms in anything other than an ordinary sense. But I certainly can't demonstrate that there's not some occult sense he's employing. I can however complain that it is occult.

on use 

Finally, what of this?:

Didn’t Wittgenstein have something to say about the meaning of a word being its use? If a word is being used by an entire community of scientists, can we not recognise that use as legitimate, even if a formal definition is lacking?

Well, I think it depends on what's meant both by 'can we' and by 'use'! 

It's certainly true that there are indeed times when words are used with bona fide meanings despite the absence of any clear definition. I see no reason why this shouldn't obtain in science as well as in everyday life or in humanities discourses etc. 

And it's surely true that we might often do well to presume that a word has sense on its users' lips. Innocent until proven guilty!, we might say, for matters of meaning.

Ludwig Wittgenstein
But it's surely not the case that use inexorably makes for meaning. Just as a whole scientific community can be in error about matters of fact, I see no reason to think that they might not sometimes also be in error regarding whether certain of their own terms are meaningful. Or, to put it otherwise, I see no reason to suppose that scientists may not sometimes be 'held captive by pictures' or, to again put it otherwise, be unwittingly subject to 'illusions of sense'. (Some sciences are, I suspect, rather more prone to such illusions of sense than others. I have in mind physics like string theory, economics, psychology and cognitive science as the major contenders... but don't plan to argue this any time soon!) Newtonian physicists talked of 'absolute motion/space/duration', but this chat of theirs didn't ipso facto make such concepts meaningful! For as Wittgenstein might have put it, Newton had already unwittingly 'sublimed the logic of the language' of motion/etc, so his terms no longer enjoyed meaning.

To understand how Wittgenstein's comments on meaning and use don't commit him to the idea that any widespread use inexorably makes for meaning, it helps to distinguish between two senses of 'use'. In the first sense 'use1' refers to however a term or phrase is deployed by those who utter or write it. It refers, that is, to the mere fact of its deployment, the mere fact that people voice it. In the second sense, 'use2' contrasts with 'misuse'. (Use1 includes both uses2 and misuses2, but also includes such uses of terms as have no meaning and so which can't even be misused.) Now, people often mistakenly think that when Wittgenstein talked of the internal relationship between meaning and use, it was use1 he had in mind - as if he were trying to provide a reductive definition of 'meaning' in terms of how we shunt words about. I think this quite wrong, and that it's only the correct uses, the uses2, of a term that are of a piece with its meaning. (Glock: 'Meaning is a matter not of how an expression is actually used and understood, but of how it is (or ought to be) used and understood by members of a linguistic community. What is semantically relevant is the correct use of expressions.') So, to summarise: whether a community's deployment of an expression is meaningful is not something guaranteed simply by their using it. And whether it's meaningful will instead amount to whether it's use can be elucidated, whether it avoids oscillating unstably between different senses encouraging the making of illicit inferences, whether clear negations of propositions deploying the term can be formulated, and sometimes, yes, even whether it can be clearly defined.